EVAPORATION AND CONDENSATION 305 



cm. of tungsten surface. As the temperature of the filament is raised the 

 rate of disappearance of nitrogen increases exactly in proportion to the 

 rate of evaporation of the tungsten. 



In a similar way it has heen found -^ that tungsten atoms combine on 

 the first collision with carbon monoxide molecules to form WCO, and that 

 platinum atoms combine on the first collision with oxygen molecules to 

 form Pt02. 



A remarkable case is that of molyl)denum and nitrogen. There is 

 definite evidence ^'^ that each collision between an atom of molybdenum 

 and a molecule of nitrogen results in combination, but there are apparently 

 two kinds of compound formed. One is a stable chemical compound which, 

 when it deposits on the bulb, cannot be decomposed by heating to 350° C, 

 while the other is an extremely unstable compound which decomposes 

 spontaneously when it strikes the bulb. The stable compound tends to be 

 formed in larger amoimt when the velocity of impact of the molybdenum 

 atom and the nitrogen molecule is small; that is when either the bulb 

 temperature or the filament temperature is low. On the other hand, a high 

 velocity of impact (high temperature of bulb or filament) favors the forma- 

 tion of the unstable substance. 



There are indications of a similar behavior when a tungsten filament 

 is heated to a very high temperature (3000° K. or more) in hydrogen at 

 low pressure, only in this case the compounds formed are less stable. 



II. THEORETICAL CONSIDERATIONS 



The phenomena of condensation, evaporation and reflection of mole- 

 cules are closely related to those of the viscosity, heat-conductivity and 

 adsorption of gases at low pressures. 



Various hypotheses regarding the mechanism of these phenomena have 

 been proposed. Maxwell ^^ in considering the surface conditions of a gas 

 in contact with a moving solid, assumed that "of every unit of area a 

 portion /, absorbs all the incident molecules, and afterwards allows them 

 to evaporate with velocities corresponding to those in still gas at the tem- 

 perature of the solid, while a portion (i — /) perfectly reflects all the 

 molecules incident upon it." 



Smoluchowski ^'^ used a similar hypothesis in regard to the temperature 

 drop near a surface in a gas at low pressure. He also suggested a second 

 hypothesis, namely. When molecules of a temperature Ti strike a surface 



^^ Jour. Amer. Chem. Soc, S7, iiS9, 1915. 

 ^"^ Jour. Amer. Chem. Soc, 37, 1157, 1915. 

 ^^ Phil. Trans., 170, 249, 1879. 



